lapack.h

Go to the documentation of this file.

/*
 * Copyright (c) 2010-2014:  G-CSC, Goethe University Frankfurt
 * Author: Martin Rupp
 * 
 * This file is part of UG4.
 * 
 * UG4 is free software: you can redistribute it and/or modify it under the
 * terms of the GNU Lesser General Public License version 3 (as published by the
 * Free Software Foundation) with the following additional attribution
 * requirements (according to LGPL/GPL v3 §7):
 * 
 * (1) The following notice must be displayed in the Appropriate Legal Notices
 * of covered and combined works: "Based on UG4 (www.ug4.org/license)".
 * 
 * (2) The following notice must be displayed at a prominent place in the
 * terminal output of covered works: "Based on UG4 (www.ug4.org/license)".
 * 
 * (3) The following bibliography is recommended for citation and must be
 * preserved in all covered files:
 * "Reiter, S., Vogel, A., Heppner, I., Rupp, M., and Wittum, G. A massively
 *   parallel geometric multigrid solver on hierarchically distributed grids.
 *   Computing and visualization in science 16, 4 (2013), 151-164"
 * "Vogel, A., Reiter, S., Rupp, M., Nägel, A., and Wittum, G. UG4 -- a novel
 *   flexible software system for simulating pde based models on high performance
 *   computers. Computing and visualization in science 16, 4 (2013), 165-179"
 * 
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
 * GNU Lesser General Public License for more details.
 */
 
#ifndef __H__UG__CPU_ALGEBRA__LAPACK_H__
#define __H__UG__CPU_ALGEBRA__LAPACK_H__
 
#include "common/common.h"
#include <complex>
 
/*
#ifdef LAPACK_AVAILABLE
#include <clapack.h>
#endif
*/
namespace ug {
 
/*#ifdef __CLPK_integer
typedef __CLPK_integer lapack_int;
typedef __CLPK_real lapack_float;
typedef __CLPK_doublereal lapack_double;
typedef __CLPK_ftnlen lapack_ftnlen;
#else */
typedef int lapack_int;
typedef float lapack_float;
typedef double lapack_double;
typedef int lapack_ftnlen;
 
//#endif
 
enum eTransposeMode
{
  ModeNoTrans = 0,
  ModeTranspose = 1,
  ModeConjTranspose = 2
};
 
extern "C"
{
  // factor system *GETRF
    void sgetrf_(lapack_int *m, lapack_int *n, lapack_float *pColMajorMatrix, lapack_int *lda, lapack_int *ipiv, lapack_int *info);
    void dgetrf_(lapack_int *m, lapack_int *n, lapack_double *pColMajorMatrix, lapack_int *lda, lapack_int *ipiv, lapack_int *info);
    void cgetrf_(lapack_int *m, lapack_int *n, std::complex<lapack_float> *pColMajorMatrix, lapack_int *lda, lapack_int *ipiv, lapack_int *info);
    void zgetrf_(lapack_int *m, lapack_int *n, std::complex<lapack_double> *pColMajorMatrix, lapack_int *lda, lapack_int *ipiv, lapack_int *info);
 
  // solve system *GETRS
    void sgetrs_(char *trans, lapack_int *n, lapack_int *nrhs, const lapack_float *pColMajorMatrix,
        lapack_int *lda, const lapack_int *ipiv, lapack_float *b, lapack_int *ldb, lapack_int *info);
    void dgetrs_(char *trans, lapack_int *n, lapack_int *nrhs, const lapack_double *pColMajorMatrix,
        lapack_int *lda, const lapack_int *ipiv, lapack_double *b, lapack_int *ldb, lapack_int *info);
    void cgetrs_(char *trans, lapack_int *n, lapack_int *nrhs, const std::complex<lapack_float> *pColMajorMatrix,
        lapack_int *lda, const lapack_int *ipiv, std::complex<lapack_float> *b, lapack_int *ldb, lapack_int *info);
  void zgetrs_(char *trans, lapack_int *n, lapack_int *nrhs, const std::complex<lapack_double> *pColMajorMatrix,
      lapack_int *lda, const lapack_int *ipiv, std::complex<lapack_double> *b, lapack_int *ldb, lapack_int *info);
  
  // invert system
  void sgetri_(lapack_int *n, lapack_float *pColMajorMatrix, lapack_int *lda, const lapack_int *ipiv, 
         lapack_float *pWork, lapack_int *worksize, lapack_int *info);
  void dgetri_(lapack_int *n, lapack_double *pColMajorMatrix, lapack_int *lda, const lapack_int *ipiv,
         lapack_double *pWork, lapack_int *worksize, lapack_int *info); 
}
 
 
inline char TransposeModeToChar(eTransposeMode t, bool isComplex)
{
  switch(t)
  {
  case ModeNoTrans:
    return 'N';
  case ModeTranspose:
    return 'T';
  case ModeConjTranspose:
    return (isComplex ? 'C' : 'T');
  default:
    UG_THROW("wrong tranpose mode");
    return 'N';
  }
}
 
 
// factor system
//--------------------
 
inline lapack_int getrf(lapack_int rows, lapack_int cols, lapack_float *pColMajorMatrix, lapack_int lda, lapack_int *pPivot)
{
    lapack_int info;
    sgetrf_(&rows, &cols, pColMajorMatrix, &lda, pPivot, &info);
    return info;
}
 
inline lapack_int getrf(lapack_int rows, lapack_int cols, lapack_double *pColMajorMatrix, lapack_int lda, lapack_int *pPivot)
{
    lapack_int info;
    dgetrf_(&rows, &cols, pColMajorMatrix, &lda, pPivot, &info);
    return info;
}
 
inline lapack_int getrf(lapack_int rows, lapack_int cols, std::complex<lapack_float> *pColMajorMatrix, lapack_int lda, lapack_int *pPivot)
{
    lapack_int info;
    cgetrf_(&rows, &cols, pColMajorMatrix, &lda, pPivot, &info);
    return info;
}
 
 
inline lapack_int getrf(lapack_int rows, lapack_int cols, std::complex<lapack_double> *pColMajorMatrix, lapack_int lda, lapack_int *pPivot)
{
    lapack_int info;
    zgetrf_(&rows, &cols, pColMajorMatrix, &lda, pPivot, &info);
    return info;
}
 
 
// solve system
//---------------
 
/*
 *  getrs solves a system of linear equations
 *     A * X = B  or  A' * X = B
 *  with a general N-by-N matrix A using the LU factorization computed
 *  by getrf.
 *
 *  \param  transposeMode
 *      Specifies the form of the system of equations:
 *          = NoTranspose :  A * X = B
 *          = Transpose:  A**T * X = B
 *          = ConjugateTranspose :  A**H * X = B
 *
 *  \param  n order of the matrix A.  N >= 0.
 *
 *  \param  nrOfRHS
 *      The number of right hand sides, i.e., the number of columns
 *          of the matrix B.  nrOfRHS >= 0.
 *
 *  \param  pColMajorMatrix the Matrix A in column major ordering.
 *      float/double/std::complex<float> or std::complex<double> array, dimension (lda,n)
 *          The factors L and U from the factorization A = P*L*U
 *          as computed by getrf.
 *
 *  \param  lda The leading dimension of the array A.  LDA >= max(1,N).
 *
 *  \param  pPivot  The pivot indices from DGETRF; for 1<=i<=N, row i of the
 *          matrix was interchanged with row pPivot[i].
 *
 *  \param  pRHS array, dimension (ldb,nrOfRHS) (col major)
 *          On entry, the right hand side matrix B.
 *          On exit, the solution matrix X.
 *
 *  \param  ldb The leading dimension of the array B.  LDB >= max(1,N).
 *
 *  \return 0:  successful exit. < 0: -i, the i-th argument had an illegal value
 *
 */
inline lapack_int getrs(eTransposeMode transposeMode, lapack_int n, lapack_int nrOfRHS, const float *pColMajorMatrix, lapack_int lda,
      const lapack_int *pPivot, lapack_float *pRHS, lapack_int ldb)
{
    lapack_int info;
    char _trans = TransposeModeToChar(transposeMode, false);
    sgetrs_(&_trans, &n, &nrOfRHS, pColMajorMatrix, &lda, pPivot, pRHS, &ldb, &info);
    return info;
}
 
inline lapack_int getrs(eTransposeMode transposeMode, lapack_int n, lapack_int nrOfRHS, const double *pColMajorMatrix, lapack_int lda,
      const lapack_int *pPivot, lapack_double *pRHS, lapack_int ldb)
{
    lapack_int info;
    char _trans = TransposeModeToChar(transposeMode, false);
  dgetrs_(&_trans, &n, &nrOfRHS, pColMajorMatrix, &lda, pPivot, pRHS, &ldb, &info);
    return info;
}
 
/*lapack_int getrs(eTransposeMode transposeMode, lapack_int n, lapack_int nrOfRHS, const std::complex<lapack_float> *pColMajorMatrix, lapack_int lda,
      const lapack_int *pPivot, const std::complex<lapack_float> *pRHS, lapack_int ldb)
{
    lapack_int info;
    char _trans = TransposeModeToChar(transposeMode, true);
    cgetrs_(&_trans, &n, &nrOfRHS, pColMajorMatrix, &lda, pPivot, pRHS, &ldb, &info);
    return info;
}
 
lapack_int getrs(eTransposeMode transposeMode, lapack_int n, lapack_int nrOfRHS, const std::complex<lapack_double> *pColMajorMatrix, lapack_int lda,
      const lapack_int *pPivot, const std::complex<lapack_double> *pRHS, lapack_int ldb)
{
    lapack_int info;
    char _trans = TransposeModeToChar(transposeMode, true);
  zgetrs_(&_trans, &n, &nrOfRHS, pColMajorMatrix, &lda, pPivot, pRHS, &ldb, &info);
    return info;
}*/
  
/*
 *  getri computes the inverse of a matrix using the LU factorization
 *  computed by getrf.
 *  This method inverts U and then computes inv(A) by solving the system
 *  inv(A) *L = inv(U) for inv(A).
 *
 *  \param  n (in)
 *          The order of the matrix A.  N >= 0.
 *
 *  \param  pColMajorMatrix (in/out) double array, dimension (LDA,N)
 *          On entry, the factors L and U from the factorization
 *          A = P *L *U as computed by getrf.
 *          On exit, if return = 0, the inverse of the original matrix A.
 *
 *  \param  lda (in)
 *          The leading dimension of the array A.  LDA >= max(1,N).
 *
 *  \param  pPivot (in) integer array, dimension n
 *          The pivot indices from getrf; for 0<=i<N, row i of the
 *          matrix was interchanged with row pPivot(i).
 *
 *  \param  pWork (workspace/output) double array, dimension max(1, worksize)
 *          On exit, if return=0, then work[0] returns the optimal worksize.
 *
 *  \param  worksize (in)
 *          The dimension of the array pWork.  worksize >= max(1, n)
 *          For optimal performance worksize >= n * nb, where nb is
 *          the optimal blocksize returned by ilaenv.
 *
 *          If worksize = -1, then a workspace query is assumed; the routine
 *          only calculates the optimal size of the pWork array, returns
 *          this value as the first entry of the pWork array, and no error
 *          message related to worksize is issued by XERBLA.
 *
 *  \return = 0:  successful exit
 *          < 0:  if return = -i, the i-th argument had an illegal value
 *          > 0:  if return = i, U(i,i) is exactly zero; the matrix is
 *                singular and its inverse could not be computed.
 */
inline lapack_int getri(lapack_int n, lapack_float *pColMajorMatrix, lapack_int lda, const int *pPivot,
         lapack_float *pWork, lapack_int worksize)
{
  int info;
  sgetri_(&n, pColMajorMatrix, &lda, pPivot, pWork, &worksize, &info);
  return info;
}
  
inline lapack_int getri(lapack_int n, lapack_double *pColMajorMatrix, lapack_int lda, const int *pPivot,
    lapack_double *pWork, lapack_int worksize)
{
  int info;
  dgetri_(&n, pColMajorMatrix, &lda, pPivot, pWork, &worksize, &info);
  return info;
}
 
}
 
 
#include "lapack_interface.h"
 
#endif // __H__UG__CPU_ALGEBRA__LAPACK_H__